The difference in population risk for major depression (MD) between men and women is one of the most reliable findings in psychiatric epidemiology. However, the determinants of sex differences in MD remain unclear and their combination into integrated etiologic models is lacking. This proposal, which is a competitive renewal of a large-scale, population based, longitudinal study of female-female, male-male and male-female twin pairs ascertained through the birth-certificate-based Virginia Twin Registry, requests funds to support further data analysis focused on understanding sex differences in risk for MD. Ten specific aims are articulated. We seek to explore sex differences in the exposure to and sensitivity to the depressogenic effects of stressful life events, in the psychopathologic pathway to MD and in the speed of kindling for MD. We will examine how social roles for men and women (acting directly or by modulating the impact of stressful life events) impact on risk for MD. We will attempt to determine if genetic factors which influence mood-response to cyclic ovarian hormones can explain some of the sex differences in genetic risk for MD. Using structural equation twin modeling, we seek to understand sex differences in the relationship between key personality traits and risk for MD and in the underlying causes of patterns of comorbidity between MD and other psychiatric and drug-use disorders. By studying male and female monozygotic twin pairs discordant for a lifetime history of MD, we will clarify sex differences in the environmentally determined correlates of MD. Finally, our collection of extensive information on proximal and distal risk factors for MD permit us to construct an integrated, developmental model for the prediction of MD and to explore the differences in such pathways in men and women. These analyses take advantage of a number of unique aspects of this sample including i) the twin design, ii) large sample size, iii) epidemiological sampling frame, iv) longitudinal assessments, v) evaluation of a wide range of common psychiatric and substance use disorders and an extensive series of risk factors and, of particular importance for this study, vi) a large sample of opposite-sex dizygotic pairs -- probably the best natural experiment for the study of sex differences in humans.